STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact email@example.com. A CRITIQUE OF PROFESSOR WIRTH'S METHODS OF MEASUREMENT OF ATTENTION By t,. R. GeiSSLBR I have been working, for the past two years, upon the problem of the measurement of the degree of attention; and I have sought to ap- proach the problem in the way that has already led to success in the case of intensity of sensation, by the assignment of numerical values to introspectively differentiated degrees of clearness. The problem has already been attacked by experimental psychologists, and notably by Professor W. Wirth, of Leipzig. Wirth's methods are entirely dif- ferent from my own: but his general formulation of the problem, as that of measurable degrees of clearness, is the same. As I have found Wirth's articles both difficult and obscure, and as my criticism of his work must be given with some fullness if it is to be itself intelligible, I have thought it well to devote a special paper to the discussion of his contributions to the subject. There is a difficulty at the outset, due to the fact that Wirth ex- presses himself only incidentally as to the relation of the field of attention to the field of consciousness. Wundt has made ns familiar with the distinction between attention and inattention, apperception and perception, inner point of regard and inner field of regard. Wirth, on the other hand, appears (although I wish to speak on this point with great caution) to identify the range of attention with the range of consciousness. He aims to give "a precise determination of the range of simultaneous mental processes." ] The title of his article in the Wundt Festschrift is "Zur Theoriedes Bewusstseinsumfanges und seiner Messung." He speaks of the "Versuch einer Wiedergabe des gesammten Bewusstseinsumfanges." 2 Phrases like these seem to show at least that he is concerned with a total simultaneous con- sciousness, with both focal and marginal processes, even if they do not in themselves show that he draws no definite distinction between centre and periphery of consciousness. But we find, later, a passage like this: "the mere determination of the possible number of (tachis- toscopically) isolated elements in maximal clearness does not at all show how high the absolute clearness-degrees may rise, or how many objects of the same clearness may be added to the average number of five isolated single objects, provided that all of them together form a familiar, associatively related whole." 8 Or again: "our measure of range really presupposes only a uniformly graded and highest possi- ble average clearness of the whole complex. 4 " Here we have, appar- ^■Pkilosophische Studien, xx, 489. *Op. tit., 493. 'Op. tit., 524. "Durcb die Feststellungder moglicben Zahl derartig (tachistosko- pisch) isolirter Elemente in maximaler Klarheit ist jedoch noch gar nichts dariiber bestimmt, wie hoch sich nun die absoluten Klarheitsgrade belaufen bezw. wieviele Objekte noch in der gleichen Klarheit wie bei ca. 5 isol irten Einzelobjekten zu diesen hinzutreten konnen. wenn sammtliche Objekte zusammen ein gelaufiges, untersich associativ verbundenes Ganzes ausmacben." *Ofi. til , 524. "Zunachst ist ja eigentlich nur die gleichmassig abgestufte und im Mittel moglichst erhohte Klarheit des gesammten Complexes zu unserer Messung des Umfanges vorausgesetzt." MEASUREMENT OF ATTENTION 121 ently, an identification of attention and consciousness. And this inference is borne out by a passage in which Wirth discusses the rela- tion between the apperceptive and perceptive regions of a simultane- ous consciousness, and declares outright that there is no sharp difference between the two, but that on the contrary there are a num- ber of transitional stages as we pass from the higherto the lower level. It is evident, he says, to unaided introspection "that the contraposi- tion of an apperceptive and a perceptive region is not to be thought of as a mere dual division of consciousness. Within the simultaneous whole, several stages of attention and of clearness may always exist side by side, according as at any moment a larger or smaller number of unitary complexes forms the immediate experience of the subject. Although under certain conditions, which favor a kind of dual divi- sion, it is possible that a region, to which a fairly uniform attention is given, may be opposed to a 'background' or 'periphery' of con- sciousness, to which attention is as uniformly denied, nevertheless the concept of apperception ordinarily denotes a general direction of the process which goes on over the whole field, though at different places with unequal completeness." 1 Apperception is thus, as it seems, simply a general tendency to conscious self-realization, com- mon to all the contents of a consciousness, but carried farther in the case of certain mental processes than in that of others simultaneously present. The antithesis of 'dunkel bewusst' and 'klar bewusst' practi- cally lapses, as antithesis, and the difference between perception and apperception becomes merely a matter of degree. If it is objected that, even in Wundt himself, this difference ha§ never been anything more, we reply that while the objection may be formally sound, in terms of a strict definition of the conscious states known as clear and obscure, nevertheless the distinction has played so important a part in the Wundtian system as to be, to all intents and purposes, a dis- tinction of kind. This generic, Wundtian difference seems to be given up by Wirth, without defence or discussion; and we may add that Wundt himself, in his brief discussion of Wirth's experiments, 2 vacil- lates between the expressions 'range of consciousness' and 'range of attention' as if he too saw no reason to distinguish them. The question at issue, however, is more than a question of personal belief or of an individual psychological system; it is a question of fact, of observation. That Wirth would have done well to face it as a ques- tion of fact will come out, I believe, in the course of the present paper. The above interpretation of Wirth's attitude is offered with all reserve, since his language — if I understand it aright — is not always consistent. I turn now to a consideration of his experimental investi- gations. Of the three sets of experiments reported 8 the first is of a prelimi- nary character, and is of no importance for the determination of degrees of clearness. In these experiments, a standard visual complex of 25 ^Op. rit., 493. 'IHierbei zeigt sich vor allem noch, dass die Gegeniiberstellung eitier apperceptiven und perceptiven Region nicht etwa bloss als eine einfache Zweitheilung des Bewusstseins zu denken ist. Innerhalb des simultanen Ganzen sind jederzeit mehrere Stufen der Beachtung und der Klarheit neben einander moalich, je nach der grBsseren Oder geringeren Zahl von Kinheitsbildungen. auf welche sich das Subject augenblicklich im umnittelbaren Erleben bezieht. Wenn auch unter Umstanden in einer Art von Zweitheilung eine ziemlich gleichmassig beachtete Region einem ahnlich gleichmassig unbeachteten 'Hintergrunde,' bezw. einer 'Peripherie' des Bewusstseins gegeniiberstehen kaim, so bezeichnet doch der Begriff der Apperception mehr eine allgemeine Richtung des Processes, welcher innerhalb des ganzen Blickfeldes an verschiedenen Stellen in ungleicher Voll- kommenheit durchgefiihrt ist." sphys. Psych., IIT. 1903, 358-360. 8 Philos. Stud., XX, 1902, 635-659; Psych. Stud., II, 1906, 30-88; A. Kastner und W. Wirth, ibid., Ill, 1907, 361-392, and IV, 1908, 139-200. 122 GBISSLBR or less small geometrical objects was observed, with maximal atten- tion, until all its elements had been clearly apprehended. During a short interval, usually of less than a second, the complex was slightly varied, in one or several different places unknown to the observer. Then it was tachistoscopically re-exposed for another fraction of a second, and theobserver was asked to say whether, and if possible where, a change had taken place. Wirth's argument is as follows : "So far as, under these variable conditions, a correct judgment of difference remains possible, so far at least does the range of the simultaneous visual consciousness extend. But the difference necessary, at any given point, to arouse this correct judgment of difference, or in other words the differential limen under these special conditions of atten- tion, is at the same time a measure of the clearness-degree which obtained at that particular point in the standard complex". 1 That is to say, if the change made at a certain place in the complex is noticed during re-exposure, then the object was in consciousness at some moment of the primary exposure; and the amount of change needed to arouse notice is inversely proportional to the degree of clearness of the mental content. The results show that, if the complex contained only 12 or 13 figures, any change in them could always be judged cor- rectly; whereas, if the complex consisted of 25 figures, many errors occurred. With a momentary original exposure of a standard com- plex of 3 to 5 figures, changes in there-exposure field were recognized with correspondingly greater or less certainty. Thus the results confirm and extend those of previous tachistoscopic experiments upon the range of attention .proper. After this preliminary investigation, Wirth comes back to the "prob- lem of the completest possible determination of the range of con- sciousness", 2 which leads him more especially to "a quantitative deter- mination of the simultaneously present clearness-degrees or degrees of consciousness of the greatest possible number of elements". 8 Nothing is said of the apperceptive and the perceptive regions; and it is only by a close examination of the experimental methods and results that we shall be able to decide whether Wirth has measured different degrees of clearness within the apperceptive region only, or whether he is dealing with the whole series of assumed transitional steps between the upper and the lower levels of consciousness. Wirth devised a special apparatus, whose main features are as follows. The field of vision for the left eye consisted of the funnel-shaped and uniformly illuminated surface of a glass cone 50 cm. in diameter at the base and 25 cm. high. The left eye was placed at the centre of the perimetrical field. The whole surface was divided into 6 concentric rings, the peripheral ring being again divided into 24 regions, the next into 20, the next into 16, the next into 12, the last into 8, and the inner disc into four regions. Of the 84 regions thus demarc- ated, 37 were selected for experimentation. Tlie arrangement is shown in the Fig. By referenceto this schema, the point of change could be easily identified. The conical surface was observed continuously, 1 Philos. Stud., XX, 593 f. "So weit sich unter den Ubrigen, bereits bekannten Bedingungen noch ein richtiges Unterschiedsbewusstsein bei einer soichen Varia- tion mit Sicherheit einstellt, so weit reicht mindestens der Umf ang des simultanen optischen Bewusstseins. Die Differenz aber, welche zur Erzielung eines soichen .TJnterschiedsbewusstseins an einer bestimmten Stelle nottawendig 1st also die Un- terschiedsschwelle unter diesen speciellen Aufmerksamkeitsbedingungen, ist zu- gleich ein Mass des Klarheitsgrades, welcher an dieser Stelle bei der Verteilung innerhalb des betreffenden Complexes herrscht." 'Psycb. Stud., II, 30 f.; "die Aufgabe einer moglichst voll&tandigen Bestimmung des sog. Bewusstseinsumfanges." 'Op.cit., 31.; "der quantitativen Bestimmung der gleichzeitig vorhandenen Klar- heitsgrade Oder Bewusstseinsgrade moglichst vieler Elemente." MEASUREMENT OF ATTENTION 123 with different distributions of attention, while by the manipulation of lenses and mirrors any one of the 37 regions could be gradually changed until its brightness was noticeably different from that of the rest. The first problem, then, was to establish the 'normal' differential linien for each region while the apex was fixated and the particular region was ob- served with maximal attention. If, for example, the brightness of the whole field of vision at a given moment was 174.5 photometric units, while the observer was attending to region no. 2, and if the additional brightness necessary for the region to become just noticeably different was 29 photometric units, and the duration of thechange was 6.75ff, then the normal limen for region no. 2 was 2 ^ x '^ = .91. The values for 174-5 each of the 37 regions were similarly determined. The next question was to find out how the normal limen was changed by varying concen- tration of attention, and how the changes could be made to express or measure degrees of clearness. To answer it, Wirth divided the normal limen of any given region by its 'distraction' limen, that is, by a limen for the same region with a prescribed direction of attention. "A com- parable value for the degrees of clearness (Beachtung) realized with a total distribution of attention over the various elements, each having its variously favorable position in the complex, may evidently be ob- tained without serious error if we divide the limen for maximal atten- tion to a known region of change by the liminal value (for the same region) resulting from the method without knowledge, that is, from the method employed for the determination of the range (of con- sciousness)". 1 Or, more explicitly: "a more delicate differential or 1 Philos. Stud., XX, 598. "Einen vergleichbaren Werth der Grade der Beachtung, welche fiir eine und die uamliche Gesammtvertheilung der Aufmerksamkeit mit 124 GEISSLBR variational limen will correspond to a higher clearness degree; and the ratio of this limen to the normal limen for the same region as known in advance and made the object of attention will offer the greatest approximation to a comparable measure of degrees of clear- ness." 1 Thus, to return to our previous example of region no. 2; if we determine its distraction limen, while attention is distributed over the whole area, we find the value?2-55M = 2.06. The degree of clear- 146 4 ness corresponding to the distraction limen 2.06 is then expressed by the ration 0.91:2.06. The clearness degrees for the other regions are determined in the same way. We thus obtain 37 different ratios, which are made comparable by reduction to the same basis of 100; e.g., in the case of region no. 2, the proportion 0.91:2.06= \ao:x re- sults in the value 226 as the clearness index for a totally distributed attention. Wirth has determined in this manner the clearness indices of all the 37 regions for 6 different distributions of attention, and has presented them in 6 different schemata of the kind shown in the Fig., each region containing a ratio which represents its clearness for a certain distribution of attention. As to the psychological inter- pretation of these schemata, he says simply: "the value of these diagrams is quite independent of our psychological interpretation, which might, e.g., be given as well in terms of purely dispositional factors." 2 Nevertheless, he certainly assumes that the numerical values of the schemata are, in some way and to some extent, definite quantitative measures of degrees of clearness. In order, then, to de- termine whether and in how far the results may be accepted as indi- cations of differing clearness-degrees, we must examine the diagrams more closely. And we begin by considering them, so far as possible, from what we conceive to be Wirlh's own point of view. Wirth gives, first, a schema of minimal normal values, 8 which he employs for the calculation of the ratios with the six different distri- butions of attention. We should expect that these values would show normal characteristics as regards their magnitude, their relative fre- quency, and their spatial distribution. Now the lowest of them is 74, the highest 173, the general average 119, and the mean variation 19.3. The most frequent values (11) occur between 91 and 99, while 26 (or almost three-fourths of the 37) occur within the limits of 74 and 116, and the 11 highest values range between 123 and 173. The most regu- lar spatial distribution is evinced by the four quadrants constituting the three innermost and complete concentric areas, and consisting of 24 regions. The left lower quadrant has the lowest average value, 89±5, and the right upper quadrant the highest, i26±7; the other two are nearly equal, ioo±7 and io2±7 for the left upper and right lower quadrants respectively. However, if we look at the extreme values in each of the five concentric areas, we find that they range, as we travel in the centrifugal direction, between 91 and 107, 74 and 137, 87 verschieden giinstigerStellung der einzelnen Uleinente zu Theil werden, gewinnt man dann offenbar ohne grossen Fehler, wenn nian den Schwellenwerth bei maxi- maler Beachtung der wissentlich variirten Stelle mit dem Schwellenwerth dividirt, der beiunwissentlichem Verfahren, also den eigentlichen Umfangsbestimmungen, gewonnen worden ist." l Psychol. fitud., II, 31. "Dem hBheren Klarbeitsgrade wird hierbei eine feinere Unterschieds- bezw. Veranderungsschwelle entsprechen, deren Verhaltniss zu der 'Normalschwelle' fur die Veranderung des namlichen ira voraus bekannten und maximal beachteten Elementes die grosste Annaherung an ein vergleichbares Mass des Klarheitsgrades bieten diirfle." *0p. ctt., 73. "Der Wert dieser Karten ist ganz unabhangig von unserer psycholo- gischen Deutung, die z. B. auch im Sinne rein dispositioneller Verhaltnisse erfolgen kSnnte." s Op. cit., 71, Fig. 8. MEASUREMENT OF ATTENTION 125 and 137, 97 and 150, and 123 and 173. The difference in the central area is the smallest (16), in the next area the largest (63), and in the other areas nearly equal (about 50). In spite of these irregularities, Wirth thinks that "there appears, after all, to be a sufficient degree of uni- formity in the minimal values "'to warrant their use as 'normal' values. He therefore proceeds to divide them by the 222 distraction values, obtained with the six different directions of attention. It is obvious, however, that, if the normal values themselves show so little regularity, the 222 ratios partly depending upon them must be still less uniform as regards magnitude, relative frequency and spatial dis- tribution. Besides, they are further affected by the fact that, while a certain part of the visual field is fixated, and another part observed with greatest concentration of attention, a liminal change of bright- ness may be expected in any possible region. Surely, such complica- tions do not promise very satisfactory results. We find, in fact, that of the 222 values the lowest is 86, the highest 216. The extremes occur in the parts attended to as well as in those distracted from. For con- venience of comparison we have arranged the figures in Tables I and II under the heading Range. The second table differs from the first in having the 9 high values above 200 eliminated. If, now, these TABLE I Attention to — Total Field of Vision Range A v. MV. Area Attended TO Range av. MV. Area Distracted from Range Av. MV. Total Field 108-284 165 3° 108-284 165 30 — — — Left Half of Field 86-181 131 23 86-181 130 24 88-178 131 20 Right Half of Field 106-186 136 17 T 10-173 136 20 106-186 136 16 Left Upper Quadrant 103-206 138 20 103-171 127 18 105-206 142 20 Left Upper Periphery 91-302 147 28 135 — — 91-302 148 28 Fixation Point 93-162 "3 15 — — — 93-162 123 15 TABLE II Attention to — Total Field op Vision Range av. MV. Area Attended to Range Av. MV, Area Distracted from Range av. MV. Total Field 108-200 151 19 IO8-2O0 151 19 — — — Left Half 86-181 131 23 86-181 130 24 88-178 131 20 Right Half 106-186 136 17 I IO-I73 I36 20 106-186 136 16 Left Upper Qu. 103-198 136 19 103-171 117 18 105-198 140 19 Left Upper Per. 91-192 141 22 135 — — 91-192 141 22 Fixation Point 93-162 123 15 — — — — 93-162 t23 15 l 0p. cit., 72. "So ist schliesslich cine hinreichende Gleichmassigkeit der Minimal- werte vorhanden." 126 GEISSLBR values are really expressions of degrees of clearness, we should expect to find a marked difference in their distributions; the low values ought to be found in the parts attended to, the high values in the other parts of the field. Or, if unknown and occasional errors or the irregularity of the normal values be allowed for, we should still expect to find the average values for the parts attended to lower than those for the rest of the field of vision. The tables show no sign of such a distribution. Wirth himself ex- plains the disappointing character of the results as follows: "the effort of the special activity to bring about the right distribution of atten- tion makes itself felt as a competitive factor. With distribution of attention over the whole field, it is probable that the lowering of the clearness values is for the most part due to this disturbing factor. But the distribution of attention to a certain region also carries with it the possibility that a higher clearness-degree may be assigned to some particular point which, under the circumstances, happens to be the object of maximal attention. [For these reasons] the arrangement of the values within the observed area gives no constant picture of a determinate clearness-relievo in the visual field under steady fixa- tion." 1 Such an admission seems, however, to be fatal to the whole investigation. If the difficulties encountered were so great as entirely to obscure the results, then the experimental method employed is not applicable to the problem in hand. The third set of experiments upon clearness-degrees is reported by A. Kastner and W. Wirth. Here recourse is had, not to just notice- able brightness differences, but to a plainly supraliminal brightness- change. The clearness-indices are expressed by the times of quickest possible reaction to the discrimination of change. Otherwise the ex- perimental arrangements remain the same as in the previous work. The following groups of experiments were made: I a, simple reaction with knowledge of the region to be changed and maximal concentra- tion upon it (to furnish a series of 'normal' reaction times correspond- ing to the normal brightness determinations in the former investiga- tion); I b, complete reaction, without knowledge, and with attention distributed over some one quadrant; I c, the same with total distribu- tion of attention; II, the same, with attention upon a region in the left upper periphery; III, selective reaction without knowledge, at- tention being totally distributed, and the reacting forefingers being assigned either to the right and left halves of the visual field, or else the one to a particular region and the other to all other regions. "It seems justifiable to assume that a proportional change in the reaction time will occur, if at all, then assuredly in connection with its depen- dence upon the degree of clearness with which the motive to reaction is apprehended." 2 The clearness-index is expressed by a number which is found by multiplying the reaction time of a given region, with a certain distribution of attention, by a fraction whose numera- tor is a constant (the average of the 37 normal values, 223.80-) and l Op. cit., 87. "Indessen begitint bereits die Anstrengung der besonderen Tatigkeit zur eigentlichen Verteilung der Aufmerksamkeit sich konkurrierend geltend zu machen. Bet der Verteilung auf das ganze Sehfeld fallt wahrscheinlich der Haupt- anteil der Erniedrigung des Klarheitswertes dieser Stoning zu. Doch tragt die Verteilung der Aufmerksamkeit auf ein Gebiet jederzeit die MBglichkeit in sich, einem beliebigen bei der Verteilung gerade maximal beachteten Punkteinen hSheren Klarheitsgrad zukommen zu lassen, als ohne diese Verteilung. Die Art der Anordung der Werte innerhalb des beachteten Gebietes lasst keine Konstanz eines bestimmten Klarheitsreliefsim Sehfeld bei festgehaltener Fixation auffinden." *Psych. Stud., Ill, 300. "Wenn irgendwo, so scheint nun gerade fur die Abhangig- keit der Reaktionszeit vom Klarheitsgrade der Motivauffassuug, die wir hier im einzelnen untersuchen wollen, eine solche proportionale Veranderung der zunachst von beliebigen Unterschieden der Raumlage modifizierten Reaktions- zeiten angenommen werden zu kbnnen." MEASUREMENT OF ATTENTION 1 27 -whose denominator is the normal average value (of 5 individual values') for a given region. Thus, if the reaction time for region no 18 is 2i&r. with total distribution of attention, while the normal aver' age reaction for the same region is 221, then its clearness-index wilj be 2l6x^^-=2i8. Hence, with total distribution of attention, 218 221 represents the degree of clearness for region no. 18, independently of its position in the visual field. The results of the normal reaction experiments are uniform. The times vary only between 200 and 245, and show a fairly regular distri- bution over the whole area. On the other hand, the results with various distributions of attention are so irregular that the authors are satisfied with the general comparison of the average values for the whole area, without, as a rule, giving due consideration to the large mean variations, or even distinguishing between the regions attended to and those not attended to. Their conclusion runs: "no matter how much the distribution of the times varies, over the whole visual field, for the different adjustments of attention, the general averages agree almost perfectly in cases of approximately equal practice; their mean is 282. This time is, therefore, in a certain sense the expression of a constant energy expended in the control of the whole perceptual field of possible motives to reaction, — an energy that, under the dif- ferent distributions of attention, manifests itself in the time of prep- aration even for varying spatial positions ; so that, on the average, a higher degree of expectation of a certain group of motives lowers cor- respondingly the preparedness of the observer for all the other possi- bilities." J It is difficult to see what relation such a conclusion can sustain to the problem of measuring either the range of consciousness or the degrees of attentional clearness. Some of the irregularities are, it is true, more or less adequately explained, as due to special circum- stances, while others are termed simply "Zufalligkeiten." 2 The latter designation seems to apply, for instance, to the case of total distribu- tion of attention, where the four regions of the most extreme concen- tric area contain not only the longest and the average reaction times of the whole series, but also the two quickest reactions of that par- ticular series. However, it is not of vital importance that we should know the reason for the numerous irregularities. The fact that they occur at all, and prevent the appearance of the expected results, is sufficient to prove the inexpediency of the method. Indeed, in their closing sentence, the authors virtually admit the inadequacy of the investigation. "Further enquiry must determine how far a more de- tailed differentiation of the co-ordinations, with more than two possi- bilities of choice, may avail to furnish a more accurate representation of the clearness-relievo within any given group of motives to reac- tion." 3 Thus far we have sought to review Wirth's investigations entirely from what we regard as his own standpoint; and our conclusion is 1 Psych. Stud., IV, 163. "So verschieden als die Verteiiung der Zeiten auf das ganze Sehfeld bei den verschiedenen Einstellungen der Aufmerksamkeit ist, so stiraraen die Gesamtmittel bei ungefahr gleichen allgemeiner Uebung fast voll- st'andig iiberein und ergeben im Mittel hier 282. Diese Zeit ist also gewissermassen der Ausdruck einer konstanlen Energie zur Beherrschung des ganzen Wahrneh- mungsbereiches moglicher Reizmotive, die bei den verschiedenen Verteilungen der Aufmerksamkeit in der Vorbereitungszeit auch in verschiedener raumlicher Aufteilung zur Geltung kommt; so dass die grBssere Erwartung der einen Motiv- gruppe die Bereitschaft fiir die Ubrigeu EventualitSten im Mittel urn eln Entspre- chendes herabsetzt." a Op. tit., 148. 8 Op. tit., 200. "Es ware nun weiterhin zu untersuchen, inwiefern eine weitere Differenzierung der Zuordnungen in einer mehrfachen Disjunktion ein genauerea Abbild des Klarheitsreliefs innerhalb eines Motivbereiches ergibt." 128 GBISSLER that he has failed to solve his problem. The difficulties of his method and the many factors that complicate his observations seem to obscure the real nature of his results and to rob them of any decisive signifi- cance. It may, however, be possible to give the results a new mean- ing by an interpretation from that standpoint which, in opposition to Wirth's, makes a sharper distinction between the apperceptive and the perceptive levels of consciousness. There can be no doubt that Wirth's experiments have dealt with certain degrees of clearness. The question is whether they covered the whole range of a simultaneous consciousness, from the clearest region of apperception to the obscurest region of perception, or whether they were merely concerned with certain relatively small differences of clearness within the level of apperception itself. Wirth assumes that his low values represent a high degree of clearness. He has to admit, however, that his high values do not necessarily repre- sent a low degree of clearness. They were due, at least in part, to distracting factors. While, therefore, the low values may possibly furnish a numerical statement of the clearness of processes on the apperceptive level, Wirth has been unable either to measure the clear- ness of processes on the lower level, or to prove the existence of his assumed transitional steps between the two levels. Let us consider once more the liminal brightness-determinations obtained with various distributions of attention. The 222 values rep- resent 100 actual ratios. Now Wirth certainly does not mean that each separate ratio measures a new degree of clearness, so that there are at least 100 stages between the two extremes of his experiments. If he had thought this, he would not have disregarded the large mean variations, and have contented himself with general averages. On the other hand, if his values do embrace the whole range of consciousness, and if they demarcate in a rough way some smaller number of inter- mediate steps, then we should expect them to fall into natural groups of greatest frequency, clustered around certain averages representing these steps, and thus to form a multimodal frequency curve. In this event, the number of such groups will indicate the number of inter- mediate steps, and the difference between their averages will be a rough measure of the size of the steps. But if we examine the total frequency distribution of the 222 values, irrespective of the various kinds of attention under which they were obtained, we find only the slightest indication of grouping between the values 90 and 160. The probable meaning of this we discuss later on. It is very significant that of the 100 different values only 2 occurred 7 times (namely 130 and 144) 1 " 6 " ( " 146) 7 " 5 " ( " *°3. no. "9> I2 5> 137. !49> and 157) 12 " 4 " 21 " 3 " 44 - 2 " Although the figures vary, as was stated above, between the limits of 86 and 302, 90% of them (or 200 out of the 222) occur between 90 and 180; and, what is more, 75% (or 165) lie between 90 and 160. That there is no multimodal distribution may be easily shown by con- structing a frequency surface, in which, e. g., groups of 10 possible values are represented in each unit of the abscissa. The same fact is brought out in the following table : MEASUREMENT OE ATTENTION 1 29 Within the limits of 81 and 90 there occur 3 cases 9i a 100 tt tt 10 tt 101 tt no tt tt 26 tt in tt 120 it a 31 tt 121 n 130 tt tt 32 a 131 tt 140 tt tt 26 tt 141 U 150 It tt 30 tt 151 tt 160 tt tt 21 tt 161 tt 170 tt tt 13 tt 171 n 180 tt tt 12 tt 181 tt 190 tt a 6 tt 191 tt 200 tt it 4 tt 201 a 30a tt tt 9 tt Thus our expectation of finding natural groups of values is not real- ized. We find instead that the 222 (minus the 9 extreme cases above 200) group themselves around one general average of 136, with 75% of them clustering closely about 122. A similar distribution is found in Wirth's reaction times, as may again be seen from the construction of a frequency surface. This fact would seem to indicate that, during most of the experiments, the observer managed to maintain a fairly constant and fairly high degree of concentration, while during the re- mainder the difficulties of which we have spoken exercised a dominant influence upon the results. Such an assumption becomes still more plausible if we compare the frequency distribution of Wirth's reaction times with those of other experimenters upon the sensorial reaction. We may select for instance the results of Alechsieff as typical, since he also took sensorial or complete reactions to visual stimuli. 1 We call attention to the curves for his 4 observers A, K, F, and S. These observers were required to concentrate their attention maximally upon the stimulus. Their "Einstellung" thus differed as widely as possible from that required by Wirth. Nevertheless the distribution of their results is strikingly similar to Wirth's ; so similar that the agreement can hardly be due to accident. It therefore seems fair to suppose that Wirth, after all, maintained a uniform concentration of attention upon the field of vision, and that his degrees of clearness — in so far as they are correctly indicated by his lower values — are only small varia- tions upon the apperceptive level. With this interpretation in mind we may come back to the slight indications of grouping between 90 and 160. Near our corrected average value of 122 there is, as has been said, a relatively large fre- quency of similar values. We notice further, small groups at no and 103, where we may perhaps assume slight rises in clearness above the common or normal level. We notice also small groups at the lower values 128, 137, 146, and 156, which in a similar way we may perhaps assume to represent slight decreases of clearness. The same thing appears on inspection of the reaction times. It is true that we' are not here dealing with simultaneity in consciousness, since the values are taken from the different distributions of attention. And if separate curves are drawn for the separate distributions, they will be found to show both fewer and less marked deflections from their general apperceptive levels (especially in the case of the brightness- values, where all values above 160 should probably be eliminated). It would plainly be wrong, then, to lay any great emphasis upon these groups, in view of the extreme complications of Wirth's experimental conditions. They are at best, only very slight, and are mentioned merely in order to do full justice to the results. 1 Philos. Stud., xvi, 1900, 1 ff. 9 130 GBISSLKR All these considerations force upon us the conclusion that Wirth's numerical values, in as far as they may be admitted to measure degrees of clearness, refer to slight clearness variations on the apperceptive level only; that he has not measured the range of a total conscious- ness; and that he has given us no evidence of the existence of transi- tional degrees or steps between the apperceptive and the perceptive level. It may be added that, so far as can be seen without repetition of the experiments, Wirth's experimental arrangements were most unfavorable to a solution of his problem. All tachistoscopic experi- ments (and the conical field is, in essentials, tachistoscopic) tend to exaggerate the difference between the two main levels of conscious- ness. Under Wirth's conditions, the whole world of extraneous lights and colors was, objectively as well as subjectively, excluded from the observer's eye and mind. The difficulties of concentration of attention upon the various parts of the visual field, with mainte- nance of a constant fixation, must have required an almost hypnotic state of attention, in which the isolated tachistoscopic objects appealed to the observer with such exclusive force that the mental processes which would normally have occupied the perceptive level were driven towards or below the li men of consciousness. This is probably Wirth's justification for their neglect. But such an extreme and almost ab- normal state of attention is, evidently, not the most favorable condi- tion for the solution of his problem, in whatever form he meant this problem to be understood. It is too far removed from the conditions of daily life, where the most natural and frequent occurrence seems to be the more or less clear apperception of a single mental process or small unitary group of processes, upon a background of other, more or less obscure but still noticeable mental processes. There are one or two other points that call for criticism. In the first place, practically the only observer, especially in the last two sets of experiments, was Wirth himself, so that there is no possibility of comparing results obtained from different individuals. The excuse offered is lack of time which, however, can hardly be accepted as valid. In the second place, comparison is the more necessary in this case since it is practically impossible to repeat the experiments any- where except in the Leipzig laboratory. And yet they must be repeated, unless we agree that the methods are so faulty and so un- promising as to spare us the necessity of repetition. This, at any rate, is clear: that Wirth has failed to solve his problem. For, if we try to interpret his numerical values from his own standpoint, we find them meaningless; they are obscured or invalidated by complicating factors. And if we look at them from our own point of view, and eliminate into the bargain all doubtful cases, they appear to disprove his assumption of the existence of transitional steps between the ap- perceptive and perceptive levels. The main reasons for his failure seem to be uncertainty and ambiguity in the formulation of the prob- lem ; the impossibility of overcoming difficulties of observation ; and the restricting conditions of his experimental arrangement, which was rather unfavorable than favorable to the type of consciousness under investigation.